Techno-Economic Optimization of Rooftop Photovoltaic Systems Using Genetic Algorithm for Government Building Kemenko 3 in Nusantara Capital City

Abstract

Rooftop photovoltaic (PV) systems installed on government buildings play an important role in supporting the energy transition and reducing carbon emissions, particularly in rapidly developing urban areas. However, conventional rooftop PV system designs are typically based on deterministic simulations that focus primarily on maximizing energy production, while technical and economic performance indicators are rarely optimized simultaneously. As a result, many installations may experience suboptimal system sizing, lower self-consumption, and higher energy costs. This study proposes a techno-economic optimization framework for an on-grid rooftop PV system installed at Government Building Kemenko 3 in the Nusantara Capital City (IKN), Indonesia. The proposed approach integrates PVsyst-based baseline simulation with a Genetic Algorithm (GA) implemented in Python to optimize key design variables, including module tilt angle, azimuth angle, and system capacity. The optimization simultaneously considers multiple performance indicators, namely annual energy production, self-consumption rate (SCR), self-sufficiency rate (SSR), and the Levelized Cost of Energy (LCOE). The optimization results indicate that the optimal configuration is achieved with a tilt angle of 1.09° and an azimuth angle of 8.58°, resulting in an installed capacity of 79.80 kW and an annual energy production of 164,114.63 kWh. The optimized system achieves an LCOE of 0.0590 USD/kWh, with an SCR of 72.76% and an SSR of 51.85%, demonstrating efficient utilization of locally generated solar energy. These results confirm that GA-based optimization can significantly improve both the technical performance and economic competitiveness of rooftop PV systems, providing a practical framework for optimizing PV deployment in government buildings in tropical regions.